What happens when it becomes cheaper and more efficient to employ robots in a factory instead of humans? Where do the jobs go? Will there be enough? Let’s take a peek at supply-side economics in an information-driven robotics age.
/READ MORE// Watch a Tesla Model S Being Born
This article originally appeared in The European Business Review on November 20, 2012.
From the 1960s up until the 1990s Western manufacturing companies were out-performed by new levels of quality, cost and efficiencies emanating from Japan. Upon realizing that this new competition from the Far East was playing to a different set of rules, the West endeavored to catch up by understanding and applying principles such as Kaizen, TQM and Lean themselves
This process took decades, caused partly by assumptions that Japan’s success could be replicated simply from copying the techniques, rather than understanding that it was more an inter-woven series of principles and beliefs. Going through the motions by sitting in Quality Circles and talking ‘zero defects’ did little until the activities were aligned with the strategies and management of the organization. Even now, many Western manufacturing companies struggle to implement these principles and achieve the desired efficiency savings.
Following on from the Japanese management revolution, Michael Hammer’s concept of ‘Reengineering the Corporation’ became very popular, which, alongside the Y2K issue, helped create worldwide demand for integrated ERP systems like SAP. Once again, different companies obtained significantly different levels of ROI from these initiatives and investments. Likewise in the early 2000s the dot.com boom saw many companies jump headfirst into this digital revolution, without fully establishing the best way for this new technology to be used.
I now believe we are about to move into the next significant period of change – this time from automated systems. Recent months have seen the introduction of a number of new inventions that when viewed individually may appear to be just interesting but disparate tools – however, all is not as it seems.
Let’s go through the different innovations in each area of the Supply Chain…
DEMAND CAPTURE AND CREATION
Any innovation that supports both marketing and operations is going to have an impact – and the introduction of intelligent vending machines like the Coca-Cola Freestyle has certainly done that. Designed by Pininfarina, the Italian car design firm, it is a transformational innovation; a game changer for the FMCG industry. First introduced 3 years ago, the Freestyle allows users to choose more than 120 soda brands or create new combinations simply with the touch of a screen, using pharmaceutical micro-dosing technology derived from the medical industry.
As arch-rival PepsiCo struggles to retain market share and Reuters describes its share price as “languishing”, Coke’s creation of a new business model via the Freestyle machine has seen the company prosper once more. It is in widespread use in America and has proven widely popular, with restaurants clamoring to have the Freestyle installed, as those outlets that have installed the machine have recorded hundreds of percentage points increases in sales.
The machine is much more than a simple dispenser; it is a data capture tool and a marketing device. It also transmits supply and demand data to Coca-Cola and to the machine owner, including which brands are sold and at what time of day. The Freestyle has also demonstrated the effectiveness of releasing new trial recipes in select markets, gathering consumer feedback and then beginning full-scale production. This is a powerful concept and moves the business ahead of its competitors in terms of its drive to be truly demand driven. A Coca-Cola spokesman stated that,
Freestyle’s data tracking technology gives us the ability to gather consumption data to optimize our product offering and assess where there are opportunities to create new retail brands. These machines also help our trade customers to manage their beverage inventory more effectively so they have the right brands in stock.
Coke Freestyle machines are GPS-enabled, supporting location management and movement; they self-diagnose, sending alerts when it encounters a problem, allowing for remote diagnostics and repair capabilities. They also self-replenish, sending an alert with ingredients are low.
Retailers are also changing the way they interact with their customers, and being more strategic about segmenting their offerings to suit the different needs.
Social media has been used as well by Coca Cola via ‘Virtual Freestyle’ Facebook apps that let fans interact with the dispenser before ever visiting one in real life. They could also request a dispenser to be added to a location near them, resulting in requests for dispensers in more than 1,500 towns and cities across the US – the ultimate Demand Driven Supply Chain.
Like Coke, Unilever has seen the opportunity from intelligent vending machines. Their ice cream brand Wall’s is rolling out a digital vending machine, called ‘Share Happy’, which uses face-reading technology and provides them with a dense set of sales data such as the type and number of items sold, the time of day that sales take place as well as the user’s age bracket, gender and ethnicity via face recognition software. This ‘in-the-field’ research is already being used for future product development, marketing and media buying activity.
Retailers are also changing the way they interact with their customers, and being more strategic about segmenting their offerings to suit the different needs. Long gone are the days of mass marketing where every customer was assumed to want the same thing. In Seoul, HomePlus, Tesco’s South Korea brand has established ‘virtual stores’ in the city’s underground subway, aimed solely at busy professionals who have little time to shop, and conveniently located on their route to and from work.
Using a virtual display of shelves of products selected specifically for this segment of customer, this channel allows the customer to scan the relevant QR code of the desired items with their HomePlus Smartphone App in order to procure them and organize home delivery. HomePlus’s App is now the #1 shopping App in Korea, with over 900,000 downloads since it launched in April 2011.
The whole concept has proven so successful amongst 20 to 30 year olds that HomePlus is expanding the concept out of the subway to more than 20 bus stops close to local university and other pedestrian areas.
BIG DATA, CLOUD COMPUTING AND APS
These new and innovative ways of capturing consumer information and demand generate huge amounts of data. According to the EMC Corporation, by the end of 2012, 2.7 zettabytes of global digital data will exist. What has become known as ‘Big Data’ is about the opportunity that today businesses can turn this huge amount of data into business value. Luckily the capability to collect and analyze these huge amounts of data is also growing exponentially, with numerous software companies creating new analytical tools designed to extract that business value, spawning a new technology sector in itself.
The rise of Cloud Computing has helped to drive this technological advancement, as the ‘Cloud’ offers resources without having to acquire knowledge of the systems that deliver it. Which is important as Big Data requires significant management, advanced technologies, new skills, huge amounts of external data and lots of data services – things which businesses would seriously question whether they need (or want) to manage when the real task is to focus on analysis to unlock the key insights and business value.
So now businesses are able to liaise with customers and consumers in different ways, extracting, storing and analyzing huge amounts of previously unavailable information in order to specifically tailor products and services – but what about the planning and replenishment of these goods and services?
In order to handle this ever increasing data volume and demand complexity, ERP software companies like SAP have developed Advanced Planning Systems (APS) that include complex demand planning and forecasting capabilities, supplier network planning, production planning, distribution planning, as well as activities such as automated procurement, replenishment, fulfillment and billing. Again, all this is now available in the Cloud.
The entire landscape can now be considered a network with nodes representing devices with varying capabilities and various functional capabilities. Planning of routes with p-to-date information from smart machines enables more detailed distribution planning, dramatically reducing costs and time to replenishment – but only for those companies with their data management and systems under control.
In his 2007 article ‘A Robot in Every Home’ Microsoft founder predicted that we were on the verge of a robotics revolution that would have a similar impact as the rise of the PC in the 1970s and 1980s. “When I talk to people involved in robots and look at the trends that are now starting to converge, I can envision a future in which robotic devices will become a nearly ubiquitous part of our day-to-day lives.”
Robotics is not new – according to the International Federation of Robotics there are now 1.1 million working robots in the world, and in car manufacture about 80% of the production is completed by machines. Most industrial robots are large, heavy, expensive one-armed machines capable of repeatedly performing a set of precise steps, such as lifting heavy objects, cutting metal or welding. Expensive to program, incapable of handling even small deviations, and so dangerous that they have to be physically separated from human workers by cages, they remain impractical to other types of manufacturing.
Baxter is the first of a new generation of smarter, more adaptive industrial robots. Baxter’s developer, Boston based Rethink Robotics, has designed the 3-feet tall, two-armed robot with a computer-screen face, animated eyes, and the capability to automatically adapt to changing environments through cameras, sensors and software that enable it to ‘see’ objects, ‘feel’ forces and ‘understand’ tasks.
So what can Baxter do? For now, Rethink Robotics founder, Dr. Rodney Brooks states that Baxter is designed for things like material handling, line loading and unloading, product inspection, light assembly, sorting and packaging – manual jobs typically done by people. Its marketing tagline is ‘Meet tomorrow’s workmate’.
Whereas traditional industrial robots perform one specific task with superhuman speed and precision, Baxter is neither particularly fast nor particularly precise. But it excels at just about any job that involves picking stuff up and putting it down somewhere else while simultaneously adapting to changes in its environment, like a misplaced part or a conveyor belt that suddenly changes speed.
Here’s the game changing aspects:
- Price: Baxter is priced at $22,000; around the same as the average salary for a US warehouse or production worker. Only Baxter can work all day and night, doesn’t get sick, require breaks or need holidays.
- Safety: Baxter can work seamlessly alongside its human counterparts, and constantly senses and adapts to what is going on in its environment.
- Intelligence: Baxter learns. To teach Baxter a new job, a human guides its arms to simulate the desired task, and presses a button to program in the pattern. If the robot does not understand, it responds with a confused expression. Equipped with sensors and other software to help it see and understand its environment, Baxter has also been programmed to apply common sense to its environment. For example, if it drops an object, it ‘knows’ it has to get another one before trying to finish the task.
- Ease of use: Rethink positions Baxter as being more like an application than a traditional industrial robot; a plug-and-play machine that small manufacturers can use without lengthy training of employees.
Brooks claims that the robot will acquire more skills as new software comes out and third parties invent functions, making it ideal for doing some of the most boring and physically tiring factory chores. Brooks likens Baxter to the iPhone stating that they both represent a sea change in computing technology, and like the iPhone Baxter will become a platform for others to develop new applications for.
PICK, PACK AND PUT-AWAY
Whereas robots have been used in manufacturing for some time, robots are also coming of age in a different area of the Supply Chain – the warehouse. Reducing the time that workers spend retrieving products from shelves for packing has long been a priority for many companies in order to be more responsive. The problem: traditional warehouses rely on fixed shelving, bad placement algorithms and error-prone humans. The result is an expensive operation with at best 95-98% operational excellence that has to scale linearly – the more you want to ship, the more warehouse space you need.
Like Dr. Brooks, Kiva founder Mick Mountz realized that there had to be a breakthrough in cost reduction for large scale distribution businesses to move to the next level.
The solution: a warehouse that took advantage of all the technological advances such as advanced algorithms, robotics and sophisticated software to tie it all together. Have the shelves come to the packing stations at the warehouse, rather than having workers go and retrieve each product from the shelves.
That simplicity belies the amazing economic power that a robotic warehouse system – including the necessary logistics, control systems, and software – can have in streamlining the mundane but vital task of getting goods where they are going. Kiva’s system uses game-changing automation technology that Kiva calls its “magic shelf,” or the Kiva Mobile-robotic Warehouse Automation System. Instead of having human workers walk through aisles of items across huge warehouse floors, robots do the moving and the workers stand still.
The average retail warehouse needs 20-40 people working a single shift. Kiva robots cut that down by 60-80%.
To achieve this, Kiva and its competitors have had to innovate in navigation, control systems and warehouse equipment. Kiva’s ‘bots’ can operate in multi-level facilities through the use of robot-operated elevators, making them practical in high-cost areas where warehouse owners need to save floor space. Empty bots are programmed to travel under the specially-designed pods when possible, leaving the aisles open not for the humans, but for other bots carrying pods. Behavioral programming techniques allow the bots to be working on several different objectives at once – traverse the floor, avoid other bots, align their guidance systems, and get recharged as needed. Parallel programming allows many different robots to simultaneously work on different parts of the same order.
All this technology has delivered three major benefits:
- Cost: The average retail warehouse needs 20-4 people working a single shift. Kiva robots cut that down by 60-80%.
- Increased throughput and capacity: Kiva’s robots actually self-organize and optimize the product shelves more optimally than in human-based fixed warehouses, leaving more capacity for storage and increased number of orders shipped.
- Decreased errors: Kiva’s software tracks every order and constantly monitors error-prone humans.
- The impact these benefits has had on one of the companies that used this technology was obviously significant – for on the 19th March 2012 Amazon announced that it had bought Kiva Systems for $775 million in cash.
It’s not just the factory and warehouses being transformed to a place where people become passive observers. On the 23rd October 2012 California Governor Jerry Brown signed a new bill that allows autonomous cars on the state’s roads. “Today we’re looking at science fiction becoming tomorrow’s reality – the self-driving car,” Brown stated.
Autonomous cars use a combination of computers, sensors and lasers to operate without the guidance of a driver. Google has modified a fleet of Toyota Prius that drive themselves using video cameras, radar sensors, laser rangefinder and detailed maps, creating a ‘virtual buffer zone’ around obstacles that makes it more aware than human drivers. The fleet has just logged over 30,000 miles without an accident through a ‘wide-range of traffic conditions’, the equivalent of 12 round-the-world trips. Google suggest the cars could come to the market within three to five years, taking the car from being simply a mode of transport to a mobile office.
Most of the major car companies all have advanced self-driving car projects in the works – for example Volvo announced that in 2014 it will offer a traffic-jam assistance system that allows its cars to automatically follow vehicles ahead of them int raffia moving at speeds up to 30 mph (48 kph). The U.S. Transportation Department in August started a field test of almost 3,000 so-called connected vehicles in Michigan. The cars are equipped with wireless devices that use global positioning systems to communicate with other vehicles and roadside systems at intersections. Eventually, it will be so clear to everyone that the computer is safer without the human driver, the truly driverless cars will be legalized.
Chris Urmson, an engineering lead for Google, said: “With each breakthrough we feel more optimistic about delivering this technology to people and dramatically improving their driving experience. We’ll see you on the road.”
But it’s not just cars.
Australia, with hits sky-high labor costs, has driverless trucks, trains and drills being used in many mines, replacing perhaps the world’s highest paid truck drivers. Multinational mining giant Rio Tinto has declared that the skills crisis and demand for greater productivity has forced them to take steps towards automated mining in Queensland, with driverless trains and trucks operating at their Western Australian operations that are actually controlled 1200km away in Perth. The fleet control system prevents collisions with other dump trucks, service vehicles, other equipment or people at the mining site, making it extremely safe and reliable.
However, it is in logistics where this technology will make most compelling financial sense. Currently a truck’s active period is dictated by laws preventing drivers from spending excessive periods behind the wheel. Automated trucks remove that restriction.
Scania, in collaboration with KTH Royal Institute of Technology in Stockholm, is developing self-driving trucks with the goal to increase accessibility and reduce energy consumption. Vehicles will communicate with each other real time, determine optimal routes via GPS, and even communicate with traffic lights so it knows whether to brake or continue. One element is ‘platooning’ where 6-8 vehicles follow a leading truck at a safe distance between each truck of about 25 meters. Running the trucks even closer together would reduce drag and lower fuel consumption by about 20%.
From a technology point of view, the self-driving car and truck is ready for wide-scale use. The only barrier is for governments to legalize them and for companies to build them.
THE AUTOMATED VALUE CHAIN
While individually these may appear to be simply interesting and innovative technological advances, the real opportunity is in aligning these together to radically transform the end-to-end Supply Chain. The following describes the potential:
Automated Value Chain
- Immediate demand signals are received via intelligent machines, including information on the customer’s preferences, location, age and background, and the machine’s current stock levels and consumption data.
- These demand signals are combined and processed by Big Data analytical tools, passed down as replenishment requests to a cloud based ERP system, automatically processed and converted into either manufacturing and/or procurement demands. Purchase requisitions are automatically converted into purchase orders and electronically dispatched to the relevant supplier.
- Suppliers will process these orders automatically on their cloud based systems, which generate replenishment demands that will be picked, packed, and loaded by warehouse robots onto an automated truck. The component supplier’s system will communicate delivery information to the manufacturer’s system, and a delivery time and warehouse slot would be confirmed automatically.
- The automated truck will deliver the component parts to the supplier, where robots will put them away until the automated planning system, responding to all of the demand and supply signals, instructs the warehouse robots to move them to the assembly line.
- Manufacturing robots like Baxter will assemble the finished product – which will then be transferred by bots to the finished goods warehouse, then picked and packed onto another automated truck that will deliver the finished products based on a pre-calculated optimal route defined by the cloud based distribution planning system that is in constant GPS contact with the trucks.
- ‘Big Data’ analytical tools will be used to provide accurate replenishment forecasts, production and distribution plans, helping to continually establish the optimal cost-to-serve models, stocking profiles and manufacturing schedules.
There will be positives and negatives from the automation revolution. Robots like Baxter and Kiva’s bots could, for the first time, bring the benefits of robotics and automation to areas of work where it never made sense before. This could help Western countries compete in the global manufacturing market against low-wage labor countries, making it more efficient for companies to re-shore and make products nearer their customers.
How important is low-cost labor when you don’t actually need labor?
Another silver lining is the possibility of improved working conditions for those remaining in robot enabled warehouses, as they are typically quieter and cooler providing more reasonable working conditions, in contrast to the conditions reported at some non-robotic Amazon warehouses.
Automating with robots also creates jobs, in refitting the facilities as well as designing and building the robots, pods, and control systems. These are certainly better quality, and require more skill, than the jobs that are eliminated, although they are not nearly as numerous.
New types of jobs will be created. Take ‘Big Data’. According to McKinsey by 2018, the United States alone could face a shortage of 140,000 to 190,000 people with deep analytical skills as well as 1.5 million managers and analysts with the know-how to use the analysis of big data to make effective decisions. Big data and cloud computing has also created a new industry sector, with numerous innovative new companies rising up, and creating employment opportunities.
There will be dramatic changes to the dynamics of the workforce, with an emphasis on acquiring talent that has a high level of design and creativity skills. There will be a massive reduction in functional, ‘I’ shaped skills, but a massive increase in demand for “T” shaped skill-sets with people who understand the customer and can shape the integrated end-to-end Value Chain to meet their needs.
This in turn will have a significant impact on old-school management styles; the command and control of teams of single function employees will disappear; replaced by empowerment, multi-functional skillets and the freedom to try things, innovate, and to make mistakes without blame.
BLUE COLLAR BLUES?
A critical issue is whether the gains of these efficiencies will simply benefit the bottom line of multinational corporations. Income inequality is a growing problem in the developed world, and one of the main factors is the lack of good paying jobs for lower-skilled workers. Manufacturing and logistics used to be a place where lower-skilled workers could got to make a decent living, and robotic automation will almost certainly mean lost jobs for an already struggling low-skill workforce.
In the Robotic Age, how will the men who don’t own the machines provide for themselves?
The Far East will not be immune either. Terry Gou, the Chairman of Foxxconn has addressed the controversial issue of poor working conditions, which led to a spate of workers committing suicide, by announcing that they will build an industrial park in Taiwan to produce 1 million robots in three years in order to replace 500,000 jobs.
The march of technology, and Moore’s law, also raises the question of when robots will simply be able to build themselves.
The battle lines between man and machine are already being drawn. Australian mining unions, outmaneuvered by the introduction of autonomous trucks and diggers, are switching from strike threats and wage demands which provide increased motivation to automate, to humanistic appeals about retaining a sense of community.
Businesses in control of their Supply Chains will be able to align and exploit these new opportunities, whereas companies lacking control will not.
In the same way Japanese car manufacturers controlled the market through their ability to perform at previously unseen levels of operational efficiency, these new developments could once again threaten to create a performance divide. Businesses in control of their Supply Chains will be able to align and exploit these new opportunities, whereas companies lacking control will not. If one company is restricted by working and driving regulations, unions, minimum wages, off-shored manufacturing and long lead times, while their competitors have invested in systems that involve none of this, then two games are being played with two sets of rules and two fundamentally different costs of playing.
The real casualties will be in complacent companies that still lack clear, cohesive strategies and integrated planning and execution capabilities. Businesses still operating in functional silos will struggle to compete with their levels of inefficiencies against proactive competitors who will develop fully integrated and systemized control towers and an automated Supply Chain and may, like the Western car manufacturers in the 70s and 80s, be left trying simply to survive.
The issue will not be simply about blue collar job losses from robotic innovations – it will be from the subsequent collapse of companies that have been complacent for decades and failed to make the opportunity to understand and control their end-to-end Supply Chains. Robots need to know what to make, pick, pack and ship – and if the company hasn’t got control of its plans, schedules and data then robots will not help.
In the same way that companies with the same ERP system get widely different results, robotics will also delight some, and disappoint others. Companies cannot exploit these innovations unless they have the basics in place; and they need to wake up quickly to this fact, realize that they simply cannot to live off past performance and prepare themselves for Supply Chain 3.0 – the automated Supply Chain.
This article was originally published The European Business Review. We were given the permission to transcribe the article from the PDF for your intellectual pleasure.